• 

  NoteCraft Ai Home Page,users can craft note from topic or exitsting unstructure note

• 

  A user can give topic name,they can get structure note based on the topic.result in the next image

• 

  here is the result for the topic input

• 

  Here is example of unstructure note

• 

  in output user got the structure note with highlight ,structure paragpraph relevant bullete points

Project Story (Hackathon Submission)

Try our Project NoteCraft Ai

🚀 About the Project

Our project, NoteCraft AI, was born from a simple frustration:
taking messy notes, long essays, or unstructured text and trying to make sense of it later.
We wanted an intelligent assistant that could not only structure raw content into clean notebooks,
but also generate a full notebook from just a topic.

💡 Inspiration

During study sessions and collaborative projects, we realized how much time is wasted
reformatting notes, organizing thoughts, and making them look professional.
The idea came: “What if AI could take care of structuring text into a ready-to-use notebook format?”

That spark became NoteCraft AI.

🛠️ How We Built It

1. Frontend (React)

   • A simple UI where users can either enter a topic or paste raw notes.
     
   • The frontend sends this content to the backend for processing.
     

2. Backend (FastAPI + LangChain + Hugging Face)

   • We built an API using FastAPI to handle requests.
     
   • Integrated LangChain with Hugging Face models to process text.
     
   • Designed a custom prompt to format outputs into notebook-style Markdown
     with titles, sections, subsections, lists, highlights, and code snippets.
     

3. Deployment

   • We containerized the backend and deployed it on Render.
     
   • Managed secrets like API keys using environment variables (.env).
     

What We Learned

• How to integrate FastAPI with LangChain pipelines.
  
• The importance of prompt engineering to get structured, consistent Markdown output.
  
• How to connect a frontend to an AI-powered backend seamlessly.
  
• Deploying full-stack apps on cloud platforms like Render.
  

Challenges

• Getting the AI to consistently output Markdown without adding extra explanations.
  
• Handling CORS issues between the frontend and backend during development.
  
• Managing API authentication securely with Hugging Face tokens.
  
• Optimizing API response time while ensuring structured formatting.
  

🏆 Impact

NoteCraft AI saves time for:

• Students → Turn messy class notes into structured study material.
  
• Professionals → Format reports, transcripts, or meeting notes instantly.
  
• Writers/Researchers → Generate notebook-style outlines from just a topic.
  

In short, we built something that helps turn chaos into clarity.

🔮 Future Improvements

In the future, we want to add:

• Multi-format Export

  • Currently, users can export their notebook in PDF format.
    
  • We plan to support exports to Word (DOCX), Notion integration, and Markdown files for easier sharing and collaboration.
    

• Notes from YouTube Videos

  • Users will be able to paste a YouTube link, and our system will transcribe the video,
    then generate a structured notebook automatically.
    

• Collaboration Features

  • Shared notebooks where teams can edit and review together in real time.
    

• Personalization

  • Allow users to choose themes, formatting styles, or predefined templates
    (e.g., research paper format, lecture notes, meeting summary).
    

• Offline Mode

  • Local export and storage for students or professionals working with limited internet connectivity.
    

Built With

• fastapi
• geminiapi
• huggingfaceapi
• langchain
• next.js
• python
• render(backendhosting)
• roocode
• typescript
• vercel(frontendhosting)